(Click image for larger version)
Posted on 05/16/2011 8:27:11 AM PDT by NCjim
You get spoiled with an aircraft that under normal circumstances will not allow you to stall the aircraft. In Alternate Law mode, the aircraft blindly follows the pilot's inputs regardless of the aircraft's flight envelope.
After AF 447, pilots have been intensely trained on how to identify and fly through pitot tube icing and loss of reliable airspeed data.
AF447 stalled but crew maintained nose-up attitudeBy David Kaminski-Morrow
French investigators have disclosed that the crew of Air France flight AF447 maintained nose-up inputs to the aircraft even after the Airbus A330 entered a stall.
The inquiry has also revealed that the pilots set engine thrust variously to go-around power and idle as they battled to rescue the jet.
In an update to the loss of the A330 over the South Atlantic two years ago the Bureau d'Enquetes et d'Analyses has detailed the last few minutes of the flight. BEA said the aircraft climbed from its cruise altitude of 35,000ft towards 38,000ft and stalled, but added that the flying pilot "maintained nose-up inputs" to the controls.
BEA confirms that the captain had left the cockpit to rest, about eight minutes before the emergency on 1 June 2009, having discussed with the relief crew possible turbulence ahead of the aircraft.
The pilots altered course slightly, about 12° to the left, and as turbulence increased they opted to reduce speed to Mach 0.8.
About 2min later the aircraft's autopilot and autothrust disengaged, and remained so for the rest of the flight. This would have put the jet into 'alternate' law, meaning it lost its angle-of-attack protection.
The aircraft began to roll to the right, and as the pilot made a nose-up left input, the A330's stall warning sounded twice - an indication that the aircraft had exceeded a critical angle-of-attack threshold.
The primary flight display on the captain's side showed a "sharp fall" in speed from 275kt to 60kt, and the aircraft's angle of attack "increased progressively" beyond 10°.
While the jet had initially been cruising at 35,000ft, investigators stated that the aircraft climbed, with a vertical speed of 7,000ft/min, heading towards 38,000ft.
The pilot made nose-down inputs as well as inputs for left and right roll. The vertical speed fell back to 700ft/min, the displayed speed "increased sharply" to 215kt, and the angle of attack reduced to 4°.
In its update the BEA said the non-flying pilot "tried several times to call the captain back".
There was another stall warning and the BEA said the stall warning sounded again. The thrust levers were positioned for take-off/go-around power but the flying pilot "maintained nose-up inputs".
Angle of attack continued to increase, it added, and the trimmable horizontal stabiliser increased from a 3° nose-up position to 13° nose-up - where it stayed for the rest of the flight.
The aircraft reached 38,000ft - its maximum altitude - with its angle of attack having increased to 16°.
AF447's captain returned to the cockpit - just 90s after the autopilot had disengaged - by which time the aircraft had started its fatal descent.
As it passed through 35,000ft the angle of attack increased to more than 40° and the A330 was descending at 10,000ft/min. Its pitch did not exceed 15°, its engine power was close to 100% of N1, and the jet oscillated with rolls of up to 40°.
"The [flying pilot] made an input on the sidestick to the left and nose-up stops, which lasted about 30s," said the BEA.
Just 20s after the captain returned to the cockpit, said the BEA, the thrust levers were set to the 'idle' position, with the engines delivering 55% of N1.
Measured angle of attack values, the BEA pointed out, are only considered valid when the measured speed is above 60kt. It said that the angle of attack, when valid, always remained above 35°.
AF447's had turned almost a three-quarter circle to the right during the emergency, and - having descended for 3min 30s - it struck the ocean surface with a ground speed of just 107kt, a nose-up pitch attitude of 16.2°, with a heading of 270°.
BEA stated that the aircraft stalled but that the inputs from the flying pilot were "mainly nose-up". It added that the engines "were operating and always responded to crew commands".
It’s the unfortunate (and unforgiving) classic:
1: Stall
2: Spin
3: Crash
4: Burn
5: Die
OMG.
Yeah, Per Le Fig original report, Airbus is exonerated.
This doesn't even count as CFIT. So hard to believe.
So they stalled at FL38, and just rode the stall all the way down? Very bizarre.
As all speed indications have the same error, none is rejected by the computers and the high speed protection will eventually engage, pitching significantly up to reduce speed (you cannot override this pitch up).The pitch-up is designed to bleed speed to avoid overspeed. That notwithstanding, does the BEA report indicate that was an issue?...Switching all 3 ADR (Air Data Reference) to OFF will force reversion to direct law (thus canceling all speed protections) and you can revert to basic pitch & power flight....
Nobody rationally "believes" they're in a stall unless they're actually in a stall. Although modern jets exhibit "benign" stall characteristics, it's not like you can be in stall and not notice it. There three identification cues for a fully developed stall:
From another Flight International article:
"France's Bureau d'Enquetes et d'Analyses confirmed there was a sudden change in displayed speeds and that there was a discrepancy between that shown on the primary flight display and that on the standby instrument system."
Since the three different pitot systems were giving at least two different readings, there would be no high speed protection.
So, uncontrolled pitch up due to overspeed protection is not an issue and the graphic at my post #85 here becomes the salient issue.
They did not have dual-engine failure concurrant to unreliable airspeed indication. SO, pitch / thrust (per the graphic) would have been the immediate watchword of the day. Given the facts, the memory items do not apply to the situation at hand (no imminent immediate threat to flight safety).
The aircraft stalled after climbing to FL380. That could only occur to due improper procedure per the checklist shown. Perhaps the PM read the wrong pitch to the PF?
Whatever the case, it was at FL380 that stall onset predicated the specific fundamentally - putatively ingrained - procedure to addres the very real imminent danger presented at that time; the procedures are mutually exclusive. Given the stall, pitch / power was untenable in that stall requires drop nose to checklist determined level, and increase airspeed. What airspeed given that IAS was unreliavble? Well, gee, when the deterrent buffet goes away it must be sufficient, eh?
Then attempt to return to level flight using pitch / power procedure as mandated by unreliable airspeed checklist. IF the stall returns, increase the pitch down maneuver by an additional increment to increase airspeed sufficient to depart from stall condition (repeat as necessary).
In theory this will ultimately rip the wings off. The only way that's plausible if severe icing has destroyed aerodynamics of the airfoil.
There's no evidence that severe icing sufficient for loss of lift occured though. I believe a reasonable explanation for pilot reaction: panic.
Muadib in Dune: "Fear is the mindkiller."
Don’t forget the aircraft was in the middle of a thunderstorm cell at the time, complicating the pilot’s ability to “feel” the aircraft.
Holy Cow!
Do modern cockpits have “old fashion” instrument backups? If the computer goes TU and the primary flight instrument computer starts feeding you bad info, can you go back to a artificial horizon guage, vertical speed indicator and a turn and bank indicator?
Sounds to me like the pilots were flying based on faulty data on the flight computer console and since they were IFR they were totally confused.
(Click image for larger version)
Yep, there they are, practically in front of the left seater’s eyeballs.
Wouldn’t even have to scan.
So the flying pilot possibly held the stick back and thought the “alpha floor” stall protection would save him? That makes no sense.
The whole setup makes no sense. This is, you would imagine, a highly-trained first officer (or possibly even captain flying as relief pilot), flying for a major airline, France’s flag carrier, with thousands of hours in the air, and he makes what sounds like a very similar mistake to what the inexperienced pilots on Colgan 3407 did?
Then again, do we know what they were seeing on their flight displays and the backup airspeed indicator/altimeter? On either the Aeroperu or Birgenair 757 crashes where the pitot-static systems were blocked, they were getting simultaneous stall and overspeed warnings and got overwhelmed. I’m not a pilot but I thought the “correct” thing to do in that case was to set a power setting and fly to a particular pitch. I wonder if you can even do that on an Airbus given how automated the throttle and flight control systems are.
}:-)4
As far as instrumentation error, the stall being triggered at a 6-degree angle of attack (AoA), suggests momentary updraft. I wonder how granular those data are.
Despite a small comment about a pilot's nose-down input given, it would seem likely that it was only briefly...
As the trimmable horizontal stabilizer spent most of its time at 13degrees nose-up. I'd bet the pilot forgot that quickly, and didn't realize how it was working against him to allow him to achieve a more normal AoA in the "direct law."
The nose apparently spent little to no time at all below the horizon, which would be required to get the AoA working in the direct law.
The kind of manueuver required to regain a 4-20 degree AoA would be counter to an institutionally-trained-for-professional pilot's intuition. No pilot likes making passengers screaming their heads off thinking they're on an extreme roller coaster instead of among the leather in Business Class, but far better that than the sudden death they actually experienced.
Unlike the "diving" into the ocean mentioned somewhere, this Airbus apparently pancaked/belly-flopped after spending over three minutes in a fairly consistent position, nose high and a gentle right turn (taken by the track recordings, despite all the mention of left roll input).
Surely the other pilots were not locked out of the flight deck, were they?
The known informaiton certainly is indicative of extreme pilot error, ignorance and likely panic.
It would seem AF's pilot training will require quite a bit of increased diligence surrounding stall recovery, and the particular pilot's instructors/check pilots need their butts figuratively reamed if not summarily canned.
HF, CFII
The stall initially occurred at (going from memory here, I might be off a couple thousand) 38,000 feet. At that altitude, it doesn't take a lot of AoA to generate a stall condition.
HF
Is it possible they thought the buffeting thunderstorm winds were the reason for the “wrong” readings on the VSI and AoA. Three minutes is an awful long time to maintain that false perception.
The problem is, even before a stall condition, getting “behind the power curve”, i.e; too slow, leads to full power, nose high descents.
When learning to fly twins, the hardest part, when you lose one and you are slow, is accepting the fact that you have to put the nose DOWN, right toward those nasty power lines, RIGHT NOW, in order to eventually climb.
The Colgan Air 3407 crash approaching Buffalo is a classic example of this.
Thanks for the link. It made more sense when I watched the video.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.